应用BP神经网络模型研究麻杏石甘汤方中麻黄对甘草酸含量的影响  

Study on the Effect of Ephdra vulgaris in Maxingshigan Decoction with BP Neural Network on Glycyrrhizic Acid Content

作　　者：武孔云[1] 梁光义[2,3] 贺祝英[3] 靳凤云[3] 孙超[4] 李星[3] 

机构地区：[1]贵阳学院,贵州贵阳550003 [2]贵州省中国科学院天然产物化学重点实验室,贵州贵阳550002 [3]贵阳中医学院,贵州贵阳550002 [4]贵州省植物园,贵州贵阳550004

出　　处：《时珍国医国药》2010年第4期775-777,共3页Lishizhen Medicine and Materia Medica Research

基　　金：国家自然科学基金(No.30460154)

摘　　要：目的考察麻杏石甘汤中不同组方配伍对甘草酸含量的影响。方法采用混料均匀设计法，固定麻杏石甘汤中的甘草量，将其他药材（麻黄、杏仁、石膏）作为可变因素，以甘草酸的测定量作为考察指标，应用BP（back propagation）神经网络模型对试验数据进行学习、预测。结果BP神经网络模型对麻杏石甘汤方中甘草酸含量的预测值可用。结论在一定范围内甘草酸含量随着麻黄用量（1．0998～11．4231g）的增大而增大；麻黄用量超过11．4231g后，方中甘草酸含量稳定在7．4442mg／g。Objective To investigate the effect of Ephdra vulgaris in Maxingshigan Decoction on glycyrrhizic acid content. Methods The glyeyrrhizie acid content in Maxingshigan Decoction was fixed with tempering uniformity , other medicinal materials （ephdra vulgaris, almond, plaster） were used as variable faetors ,and the content of glyeyrrhizie acid was used as inspection tar- get, tested data were studied and forcasted with back propagation neural network model. Results The forecasted and measured value of glycyrrhizic acid contents were the same as BP neural network model. The forecasted value in Maxingshigan Decoction was available with BP neural network model. Conclusion The content of glyeyrrhizic acid augmented when ephdra content （ 1. 099 8 - 11. 4231 g） was increased within limits. When the ephdra vulgaris content exceeded 11. 433 1 g, glycyrrhizie acid content was 7. 444 2 mg/g steady.

关 键 词：麻杏石甘汤 混料均匀设计 甘草酸 BP神经网络 

分 类 号：R2-03[医药卫生—中医学]